Browsed by
Category: Software

Bluetooth Low Energy STM32WB55 Eval Kit Unboxing

Bluetooth Low Energy STM32WB55 Eval Kit Unboxing

I’ve been working on a Bluetooth Low Energy (BLE) based project for the past year, using the Nordic nRF52 integrated SoC. Its powered by a 64MHz ARM Cortex M4 core and Nordic provides good software stack and SDK support.

In keeping up with current industry developments, I was excited to hear about a month ago that STMicroelectronics announced their BLE SoC kit: the STM32WB.

I’ve also worked with ST microcontrollers for the past several years, and they also provide good chip level driver library support. The eye-catcher on their BLE chip is that it’s a dual-core design: A 64MHz ARM Cortex M4 to run the application, paired with a 32MHz Cortex M0+ core to run the wireless stack.

After a few weeks of scoping limited inventory, I was able to score an Eval kit:

One of the nice options in working with ARM microcontrollers is that we’re spoiled for choice on IDE options. The Free option is the one I featured today: GNU MCU Eclipse. It is a bit of a setup/install process, but the guide is pretty straightforward and it gets you a $0 development environment that is not feature or code-size limited.

A robot car kit to satisfy curiosity

A robot car kit to satisfy curiosity

Ever since our little ones got hooked on watching NASA/JPL Curiosity landing videos, they’ve developed a strong interest in robotics (yay!!!).

And as luck would have it, our kids would often ask me (the dad) to “build a robot”.  When the kids were younger, building models with LEGO would suffice.  But it soon became apparent they wanted a “real” machine.  I had given a design try a while back.  I spent a few evenings looking up design ideas, drafting up sketches of mechanical pieces that we could fab from hobby plywood… but never got further from the computer screen.

Historically, I haven’t been a huge fan of turnkey kits.  I had saw them to be either of low quality or charged a significant premium for the convenience of all pre-selected parts in one box.

But then I found this Elegoo Smart Robot Car Kit on Amazon:

The kit comes with an Arduino control board, an I/O board, a ultrasonic module, a bluetooth module, line following module, motors, wheels, chassis, and all hardware needed for assembling the kit.


My Value Justification:

At time of writing, this off-the-shelf kit will get to your door for $69.99 USD.  You can easily spend half that cost just on machining the chassis plates yourself.  You could alternatively individually purchase a standard Arduino board, and other modules, but you’d still have to figure how to mechanically and electrically link them together.  The board designs may be open source and be a good EDA design exercise, but good luck getting the PCBs made for < $100 before you even add in the parts BOM.

The Build:

It went really well.  The kit came with clear illustrated instructions which our 2nd grader and Kindergartner were able to follow along with just a little hand dexterity help from me for some of the tighter screw assemblies and finer connections.

Parts quality is good.  The pieces appear to be machined well.  The components and hardware aligned up very easily with little fuss.

The kids now have a robot:

The kit includes demo and learning firmware projects to load via the Arduino IDE.  We are able to upload the “auto_go” firmware to the kit and watch the car drive around on the floor.

I’m quite impressed at this kit, what it comes with, and what it enables.  The car robot as-kitted presents many learning applications, but it also lends itself to expansion and modding.

In terms of inspiring learning pursuits, robotics engages the young hungry minds in mechanical engineering, electronics design, embedded firmware development, and applications development.



Introducing Computer Science with micro:bIt

Introducing Computer Science with micro:bIt

Throughout the school year we introduce computer science concepts as appropriate for age and skill level.  Towards the end of last school year we started using the micro:bit platform as a fun learning tool with much success and plan to continue in the following weeks.

The micro:bit (available at is a plug-and-go easy to use introductory electronics and programming platform with a 5×5 LED “display” matrix, a couple buttons, and a 3-axis accelerometer.

The kit comes with a battery pack and a short micro-USB cable for power/programming.  The longer hi-visibility red cable that I used is this one:

The programming environment is web based, so any computer with a modern browser and an available USB port should work.  Our kids were able to do the labs with our relatively low-power, long battery life Samsung Chromebook 3:



Time-Lapse capture with webcam and Linux scripting.

Time-Lapse capture with webcam and Linux scripting.

Time-lapse photography is a concept of taking snapshots at a particular interval over time to record events in a low resource manner, compared to capturing a video stream.  Here, we discuss how to use a relatively inexpensive web camera and a computer to build a low-budget time-lapse platform.

Action Summary

Video capture of a video capture

Pretty much any relatively modern PC with decent storage and functional USB 2.0 ports should suffice.  I had an old big-box special on hand.

Logitech webcams tend to plug-and-play easily with Linux distros:

The Logitech C615 is the model featured.  It does 1080p capture, with some nudging.

*Even though I held up an external HDD in the footage, I didn’t get to show setting that up yet and instead relying on internal storage for now.

The “cheese” program is a quick way to test a webcam. (  To install on a Ubuntu/Mint system, use apt-get in a terminal window:

sudo apt-get install cheese

Multi-adjustable Gorilla Pod Tripod makes for camera positioning in tight corners.  Mine came as a kit with a flashlight, but here’s a standalone model:

We looked at how to use the fswebcam Command Line utility to capture images via the terminal.  You may either git clone from the project page, or on an Ubuntu/Mint system, enter in a terminal window:

sudo apt-get install fswebcam

It is a bit of trial-and-error figuring the optimal parameters for the fswebcam utility with a camera.

Then we watched how bad I am at live shell scripting, though I freely admit its not my everyday gig and that I make liberal use of internet search the few times I write shell scripts.  Can’t go wrong with an O’Reilly book on Shell Scripting if you want an at arms reach reference to thumb through:


Hope you found this useful and informative.  Having a setup like this opens up a lot of other potential applications to build upon.


Disclaimer: from a technical and functional perspective, this post is about using extending usage of commodity hardware using computer scripting.  There are some potential premise security applications for such a setup.  While anyone is welcome to use the information herein, the reader is responsible for determine appropriate for use, deployment, and maintenance of such applications for their security needs.  I/We do not explicitly endorse this type of setup as a part of a security protocol.


Flowcharts – the beginning, end, and a decision.

Flowcharts – the beginning, end, and a decision.

In software design and documentation, flowcharts are drawn to draft out algorithmic flow.  Generally this is done sometime before writing out any code to set an initial plan, but sometimes not.  More often in practical applications, flowcharts make their way into design specifications documents that outline how the software was well, designed.

A flow chart can be summarized as an illustration of a process with decisions and activities to be performed based on those decisions.

A flow chart, typically starts with a “start” terminal.

A flow is indicated by an arrow.  Arrows connect blocks together in a flow diagram

A decision to be made is written inside a diamond:

The condition for the decision is written inside the diamond.

An arrow will go into the decision diamond from the top, and branch off to the sides depending on whether result is TRUE or FALSE.

An action is written inside a rectangle

And finally, a process or flowchart ends with an “end” terminal.

To review our mini-lesson, let’s do a LAB 1: Let’s follow a Process”.

Materials needed.

  • Counting Blocks.  –I originally wrote this for schooling where we have lots of nice manipulatives, which is just fancy multi-sensory speak for things that you can count with.  Whatever physical or mental objects you can count with will suffice.

Let’s read the following process:


  1. Follow the process with 2 counting blocks.  What action did you do?
  2. Follow the process with 3 counting blocks. What action did you do?
  3. Follow the process with 4 counting blocks. What action did you do?


Further Reading

Fundamental Basics: Numerical Systems

Fundamental Basics: Numerical Systems

There’s an old joke: “There are only 10 types of people in the world: those who understand binary, and those who don’t.”  If you don’t get it now, hopefully you will after this post.

For most everyday tasks, we’ve been accustomed to using base-10, or “decimal” numerical system.  Its what we’re taught from grade school and makes the most logical sense for humans.  A quick recap on integer base-10 goes back to “ones”, “tens”, “hundreds”, and we say digits go from 0-9.

Let’s take a sample number 325.  We say 325 is comprised of (3*100) + (2*10) + (5*1):


Or, an equivalent way of notation is:

 325 case you’ve forgotten algebra, I’ve linked Google Calculator above.  For each number column, the digit represents that many 10^x.  Like, 3 in the “hundreds column” is so because 10^2 is 100, and there’s 3 of them.

It’s called base-10, because the base number is 10.

In Binary, the base number is 2, a.k.a. base-2.  In base-2, the digits go from 0-1, and each column is 2^x.  So, if we take binary “1011”:


Reading the binary digits left to right, there’s 1 of (2^3) + 0 of (2^2) +1 of (2^1) + 1 of (2^0), OR (1*8) + (0*4) + (1*2) + (1*1) = 11 decimal.

While human brains process and store numbers in base-10, computer systems store data in binary, hence our discussion as a fundamental concept for software and embedded systems.

In nomenclature, each binary digit is called a “bit”.  (binary 1011 is a 4 bit number).  Data variables and CPU architectures are often referred to as number of bits.  A 8-bit processor indicates that it natively processes data in 8-bit wide registers.  A 16 bit data variable means it can hold up to 16 bits for that variable.

||        ...
||               |--2^1
||               ||---2^0
1010 1100 0011 1011

Writing out just 16 bits already starts to look like writing out credit card numbers.  Imagine how long and tedious it gets to write out/read 32 bits, 64…

Luckily, there’s another numerical format: Hexadecimal, or “Hex”.  Hex is base-16.  There’s 16 digits.  The first ten digits are 0-9.  To avoid ambiguity, the eleventh through sixteenth digits are denoted A-F, so while binary digits go [0, 1] and decimal digits go [0, 1, 2, 3, 4, 5, 6, 7, 8, 9], hex digits go [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, A, B, C, D, E, F].

Perhaps a table will help:

|decimal | binary | hex |
|      0 |  0000  |   0 |
|      1 |  0001  |   1 |
|      2 |  0010  |   2 |
|      3 |  0011  |   3 |
|      4 |  0100  |   4 |
|      5 |  0101  |   5 |
|      6 |  0110  |   6 |
|      7 |  0111  |   7 |
|      8 |  1000  |   8 |
|      9 |  1001  |   9 |
|     10 |  1010  |   A |
|     11 |  1011  |   B |
|     12 |  1100  |   C |
|     13 |  1101  |   D |
|     14 |  1110  |   E |
|     15 |  1111  |   F |

From the table, each hex digit can be represented by a 4-bit binary value.  That’s convenient in that you can group a binary representation into 4-bits and convert it to hex:

binary: 1010 1100 0011 1011
hex:       A    C    3    B

The binary value 1010110000111011 (commonly written as 0b1010110000111011 or 1010110000111011b) is equivalent to hex value AC3B (commonly written as 0xAC3B, or AC3Bh)

To convert to decimal, the value 0xAC3B can be viewed as:


where the number has

("A" *16^3) + ("C" *16^2) + (3 *16^1) + ("B" *16^0), or 
(10  *4096) + (12  * 256) + (3 *  16) + (11  *   1)  == 44091 dec.

As with base-10 math, there are calculators available that aid in converting numerical representation.

On Windows, the “calc.exe” in programmer mode (View->Programmer) is handy:


You can also try the Google:

Screenshot - 07222016 - 10:39:37 PM Screenshot - 07222016 - 10:40:06 PM Screenshot - 07222016 - 10:40:21 PM

As a building block in software and embedded systems development, having a working knowledge of these three numerical systems are a prerequisite.  There is another format, Octal, that’s base-8, that is usually also introduced in Computer Science.  I think after going through binary and hex it isn’t to hard to figure out.